Comparison of cytopathic changes in oral hairy leukoplakia with in situ hybridization for EBV DNA

It has been observed that the cytopathic changes in hairy leukoplakia (HL) correlate with ultrastructural evidence of intra-keratinocyte herpes-type viral particles. In situ hybridization is considered to be the definitive confirmation of Epstein-Barr virus (EBV)-induced HL. This study evaluated the consistency of histopathological findings, which many believe to be diagnostic, with in situ hybridization for EBV-DNA in 60 patients with lesions clinically suggestive of HL.

Hematoxylin and eosin (H&E)-stained sections were reviewed independently by three oral pathologists who did not know the hybridization results. The presence in keratinocytes of nuclear inclusions and/or homogenization, believed to be specific for EBV in these lesions, was used as an indicator for infection. Cytoplasmic changes were evaluated separately.

With in situ hybridization, 48 cases were positive and 12 were negative. When the two methods were compared, pathologist concurrence ranged from 83% to 92%. False negatives ranged from 6% to 19%, and false positives ranged from 8% to 25%. Cytoplasmic ballooning, homogenization, and perinuclear clearing were evident in all cases of hybridization-confirmed HL; however, these changes were also noted in 75% (9/12) of the cases with negative hybridization results. Most confirmed HL cases exhibited both nuclear homogenization and inclusions, although the former was more consistently seen.

Cytoplasmic changes did not agree well with EBV-DNA hybridization results, whereas nuclear changes demonstrated good, but not complete, agreement. In appropriate clinical settings, the finding of nuclear inclusions and/or homogenization may be of diagnostic value. However, because the potential for false positives and negatives is high, H&E cytopathology should not be used as a substitute for in situ hybridization in the definitive diagnosis of oral hairy leukoplakia.